|
The FPGA Summit focuses on the use of FPGAs (field-programmable gate arrays). FPGAs are semiconductor devices containing logic elements and interconnects.
Programming an FPGA typically involves selecting logic element functions and making or breaking connections among the elements to implement a specific design. That is, programming the FPGA involves selecting active connections much like building something by connecting pieces from a set of Legos.
A single FPGA can perform many functions depending on how it is programmed, much as someone can build many objects from a single set of Legos, depending on how one connects them.
The latest FPGAs contain many thousands of logic elements plus such components as memories, serial interfaces, and even full-fledged processors.
FPGAs provide an increasingly attractive alternative to custom chips, since they offer a single-chip solution without requiring a manufacturing process (or a foundry). They are also easy to change, correct, or upgrade, and can even be reconfigured in the field. They are particularly useful for low-volume applications, prototypes, situations where requirements or specifications may change, signal interfaces, and applications where board space is at a premium. Large FPGAs can actually replace processors in applications such as signal processing that require maximum throughput (that is, hardware rather than software speeds).
|